INDEX

Explanations

words related to objects or concepts associated with specific tasks or environments

oai_token-act-pair · gpt-3.5-turbo

terms related to music, sound, or performance elements

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

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

enegger

-0.81

ADRA

-0.70

 Worldwide

-0.69

 Marketable

-0.68

itored

-0.67

HCR

-0.66

¿½

-0.62

İĭ

-0.61

 Alive

-0.61

 Instit

-0.60

POSITIVE LOGITS

coin

0.81

stack

0.73

box

0.72

bot

0.69

 shader

0.69

meter

0.65

 pedal

0.65

age

0.64

hole

0.64

pack

0.63

Activations Density 0.661%

words related to objects or concepts associated with specific tasks or environments

terms related to music, sound, or performance elements

No Comments

No Known Activations