INDEX

Explanations

mentions of the word “human”

oai_token-act-pair · gpt-3.5-turbo

references to the human condition and human experiences

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 2-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.2.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.2.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

 Transcript

-0.87

abb

-0.77

forth

-0.74

Reloaded

-0.70

INO

-0.69

REP

-0.68

arella

-0.67

 Christensen

-0.66

OHN

-0.65

armac

-0.65

POSITIVE LOGITS

 beings

1.22

 readable

0.97

 rights

0.94

istic

0.93

ized

0.88

itar

0.88

rights

0.83

 fingert

0.83

 embryonic

0.83

izes

0.81

Activations Density 0.024%

mentions of the word “human”

references to the human condition and human experiences

No Comments

No Known Activations