INDEX

Explanations

questions

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

questions ending with question marks

oai_token-act-pair · gpt-3.5-turbo

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.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

anz

-0.68

apers

-0.67

 pace

-0.67

 locks

-0.66

 towels

-0.65

 limbs

-0.64

 limb

-0.63

 tentacles

-0.63

 breeze

-0.63

 showers

-0.62

POSITIVE LOGITS

 Answer

1.00

 inquired

1.00

 Asked

0.92

 asks

0.88

 Inqu

0.87

 asked

0.86

 Questions

0.84

Huh

0.83

 rhet

0.81

#$

0.81

Activations Density 0.015%

questions

questions ending with question marks

No Comments

No Known Activations