INDEX

Explanations

specific formatting patterns in news articles, such as author names followed by specific symbols

oai_token-act-pair · gpt-3.5-turbo

indicative phrases or components related to citations or references in articles

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

 charm

-0.67

 endeav

-0.66

 Quart

-0.65

 charms

-0.60

etta

-0.59

twe

-0.57

 professionally

-0.57

bomb

-0.57

rox

-0.57

lc

-0.56

POSITIVE LOGITS

Related

0.80

 Related

0.76

eve

0.72

EVA

0.71

Recent

0.70

Washington

0.68

igham

0.67

SPONSORED

0.66

BBC

0.66

 Âł Âł

0.66

Activations Density 0.173%

specific formatting patterns in news articles, such as author names followed by specific symbols

indicative phrases or components related to citations or references in articles

No Comments

No Known Activations