INDEX

Explanations

references to the District of Columbia

oai_token-act-pair · gpt-3.5-turbo

references to the District of Columbia

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

ãĥ¼ãĤ¯

-0.93

ãĤ¯

-0.85

ãĤŃ

-0.84

ãĤ¬

-0.82

ãĥĢ

-0.81

âķĲâķĲ

-0.80

ãĤ¤

-0.79

nings

-0.78

ãĤ´

-0.78

Frames

-0.72

POSITIVE LOGITS

 Territories

0.91

 Circuit

0.79

 Heights

0.78

rimination

0.73

 Geographic

0.72

 Territory

0.72

 Telecommunications

0.71

 Canal

0.71

 Presbyterian

0.71

 Railroad

0.70

Activations Density 0.032%

references to the District of Columbia

references to the District of Columbia

No Comments

No Known Activations