INDEX

Explanations

words related to automatic actions or mechanisms

oai_token-act-pair · gpt-3.5-turbo

words related to automatic or systematic processes

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

Sean

-0.74

RAW

-0.71

nan

-0.68

ngth

-0.68

forts

-0.68

DJs

-0.67

part

-0.67

âľ

-0.66

skill

-0.64

eded

-0.64

POSITIVE LOGITS

omatic

1.33

osate

0.82

tarian

0.80

 conclud

0.77

obile

0.75

otine

0.75

 concess

0.75

phrine

0.72

 veter

0.71

uates

0.71

Activations Density 0.008%

words related to automatic actions or mechanisms

words related to automatic or systematic processes

No Comments

No Known Activations