INDEX

Explanations

questions starting with "Are" and capital letters

oai_token-act-pair · gpt-3.5-turbo

interrogative phrases or questions

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

odder

-0.94

FTWARE

-0.76

rouse

-0.75

matter

-0.69

analysis

-0.68

uncture

-0.67

ication

-0.67

isine

-0.66

 Dragonbound

-0.65

ruption

-0.64

POSITIVE LOGITS

wolves

0.95

 these

0.90

you

0.85

 those

0.82

 THESE

0.82

pas

0.81

YOU

0.77

 they

0.77

we

0.77

 there

0.75

Activations Density 0.067%

questions starting with "Are" and capital letters

interrogative phrases or questions

No Comments

No Known Activations