INDEX

Explanations

keywords related to documenting information or evidence

oai_token-act-pair · gpt-3.5-turbo

instances of the word "documented."

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

enta

-0.81

awar

-0.81

bit

-0.77

isoft

-0.76

govtrack

-0.74

nir

-0.72

neg

-0.71

adding

-0.70

committee

-0.69

atern

-0.68

POSITIVE LOGITS

 documenting

0.87

 documented

0.86

ocument

0.86

documented

0.86

urally

0.80

 feats

0.74

 documentation

0.72

 Koen

0.67

URES

0.66

 chronological

0.65

Activations Density 0.021%

keywords related to documenting information or evidence

instances of the word "documented."

No Comments

No Known Activations