INDEX

Explanations

words related to claims of responsibility in various contexts, especially pertaining to conflicts or incidents

oai_token-act-pair · gpt-3.5-turbo

references to accountability and responsibility in various contexts

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.hook_resid_pre

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

24,576

Data Type

torch.float32

Hook Point

blocks.5.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

ophon

-0.69

 expectations

-0.64

tle

-0.64

 roller

-0.64

chan

-0.63

eeks

-0.61

 Expect

-0.60

Wo

-0.59

eland

-0.59

 convol

-0.58

POSITIVE LOGITS

 culp

0.81

eous

0.81

ItemTracker

0.76

 shielding

0.75

ignty

0.75

alties

0.75

 wrongdoing

0.72

displayText

0.71

fired

0.68

bard

0.68

Activations Density 0.028%

words related to claims of responsibility in various contexts, especially pertaining to conflicts or incidents

references to accountability and responsibility in various contexts

No Comments

No Known Activations