INDEX

Explanations

names of people or places

oai_token-act-pair · gpt-3.5-turbo

proper nouns and character names in media contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 9-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.9.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.9.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

 Adin

-0.88

®

-0.78

ivan

-0.75

ohn

-0.71

obal

-0.70

orial

-0.69

VID

-0.69

eatures

-0.69

VERT

-0.68

 horm

-0.67

POSITIVE LOGITS

 indefinitely

0.70

 overhead

0.67

 following

0.66

 operations

0.64

 reverber

0.63

 expansion

0.62

 tomorrow

0.61

 fiasco

0.61

 ahead

0.61

 amidst

0.60

Activations Density 0.639%

names of people or places

proper nouns and character names in media contexts

No Comments

No Known Activations