INDEX

Explanations

specific targets or goals mentioned in the text

oai_token-act-pair · gpt-3.5-turbo

instances of the word "target" 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

ansk

-0.75

GGGG

-0.69

 Geological

-0.66

IGH

-0.66

lycer

-0.65

fo

-0.65

Created

-0.64

aucas

-0.64

ISTORY

-0.63

aug

-0.63

POSITIVE LOGITS

ted

1.18

izen

0.89

 targets

0.84

 target

0.82

ting

0.75

topic

0.75

oided

0.72

ishes

0.71

 audience

0.70

 range

0.70

Activations Density 0.020%

specific targets or goals mentioned in the text

instances of the word "target" in various contexts

No Comments

No Known Activations