INDEX

Explanations

words related to organizations dealing with research and policy, specifically think tanks

oai_token-act-pair · gpt-3.5-turbo

references to think tanks

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

VID

-0.73

âĸ¬

-0.66

 Fram

-0.65

 Chains

-0.65

Lod

-0.65

 Plain

-0.63

juries

-0.62

Words

-0.61

 dysfunctional

-0.60

 powd

-0.58

POSITIVE LOGITS

tank

0.93

aimon

0.92

anyahu

0.91

arian

0.89

eus

0.89

arians

0.88

ament

0.88

rition

0.85

erness

0.82

iness

0.81

Activations Density 0.005%

words related to organizations dealing with research and policy, specifically think tanks

references to think tanks

No Comments

No Known Activations