INDEX

Explanations

text related to identifying or describing problems

oai_token-act-pair · gpt-3.5-turbo

references to problems or issues

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

rib

-0.88

urses

-0.86

orks

-0.79

vez

-0.76

theless

-0.76

ron

-0.75

arnaev

-0.75

alty

-0.74

rica

-0.73

ensions

-0.72

POSITIVE LOGITS

 plag

1.21

 solved

1.05

 solving

1.03

atical

0.88

Problem

0.87

 confronting

0.82

atic

0.80

atics

0.80

 endemic

0.80

 facing

0.79

Activations Density 0.082%

text related to identifying or describing problems

references to problems or issues

No Comments

No Known Activations