INDEX

Explanations

proper nouns related to various locations and people

oai_token-act-pair · gpt-3.5-turbo

references to individuals and their actions or roles

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

 Unle

-0.68

scription

-0.67

raft

-0.65

 Speedway

-0.64

yss

-0.62

ongh

-0.62

 Fargo

-0.61

 Prompt

-0.61

POSITIVE LOGITS

 resided

0.87

'd

0.78

 penetrated

0.77

 boarded

0.75

 reside

0.74

 invaded

0.73

 interacted

0.73

 lived

0.72

 resides

0.72

 poked

0.71

Activations Density 0.163%

proper nouns related to various locations and people

references to individuals and their actions or roles

No Comments

No Known Activations