INDEX

Explanations

phrases relating to things being accessible or available to the public

oai_token-act-pair · gpt-3.5-turbo

phrases indicating accessibility to the public

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 dearly

-0.68

tein

-0.67

 doomed

-0.63

Dur

-0.62

 à¤

-0.62

Ð¼

-0.61

å¦

-0.59

lycer

-0.58

sis

-0.58

 attributed

-0.57

POSITIVE LOGITS

 outsiders

1.07

 everyone

0.92

 everybody

0.89

 anyone

0.88

 scrutiny

0.86

everyone

0.85

 inspection

0.85

 explore

0.85

 anybody

0.84

 public

0.83

Activations Density 0.135%

phrases relating to things being accessible or available to the public

phrases indicating accessibility to the public

No Comments

No Known Activations