INDEX

Explanations

words related to various problems and issues

oai_token-act-pair · gpt-3.5-turbo

repeated mentions of the word "problem" and its variations, indicating discussions about issues or challenges

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

rib

-0.75

arnaev

-0.70

urses

-0.69

orks

-0.69

ritch

-0.68

arag

-0.68

theless

-0.67

boro

-0.66

eele

-0.65

rary

-0.64

POSITIVE LOGITS

 plag

1.21

 solved

1.21

 solving

1.14

atical

1.05

atic

1.01

atics

0.98

atically

0.92

 facing

0.91

 confronting

0.90

 endemic

0.87

Activations Density 0.057%

words related to various problems and issues

repeated mentions of the word "problem" and its variations, indicating discussions about issues or challenges

No Comments

No Known Activations