INDEX

Explanations

phrases indicating various unspecified, but potentially impactful, outcomes

oai_token-act-pair · gpt-3.5-turbo

phrases that indicate a variety of unspecified subjects or topics

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

eka

-0.79

edia

-0.73

ESE

-0.67

ħĭ

-0.65

uca

-0.65

rate

-0.64

 Sands

-0.64

essors

-0.64

ctor

-0.63

arest

-0.63

POSITIVE LOGITS

 goodies

0.89

 configurations

0.87

 kindred

0.84

 neat

0.84

 kinds

0.84

 combinations

0.84

 imaginable

0.83

 conceivable

0.81

 fancy

0.81

 different

0.79

Activations Density 0.068%

phrases indicating various unspecified, but potentially impactful, outcomes

phrases that indicate a variety of unspecified subjects or topics

No Comments

No Known Activations