INDEX

Explanations

words related to legal terms and actions

oai_token-act-pair · gpt-3.5-turbo

prefixes and suffixes in words

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.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

Rated

-0.67

RAW

-0.66

igun

-0.64

 tremend

-0.63

icka

-0.61

iq

-0.59

Dog

-0.58

uma

-0.58

iazep

-0.58

 Tags

-0.57

POSITIVE LOGITS

selves

1.18

forth

1.03

entimes

0.97

theless

0.96

withstanding

0.92

etheless

0.81

forward

0.80

terday

0.78

 hindsight

0.71

ishly

0.71

Activations Density 0.197%

words related to legal terms and actions

prefixes and suffixes in words

No Comments

No Known Activations