INDEX

Explanations

proper nouns or specific names within a variety of contexts in text

oai_token-act-pair · gpt-3.5-turbo

instances of the word "found" and related terms indicating discovery or recognition

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 7-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.7.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.7.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

charge

-0.74

 endanger

-0.71

selling

-0.66

esc

-0.66

sis

-0.64

cele

-0.61

visory

-0.61

 punish

-0.60

 Rosenstein

-0.60

elfth

-0.60

POSITIVE LOGITS

 similarities

0.90

 inconsistencies

0.84

 numerous

0.80

 discrepancies

0.79

 anomalies

0.76

 irregularities

0.75

WER

0.75

 insights

0.74

 myriad

0.74

":[{"

0.72

Activations Density 0.238%

proper nouns or specific names within a variety of contexts in text

instances of the word "found" and related terms indicating discovery or recognition

No Comments

No Known Activations