INDEX

Explanations

links and references for further reading

oai_token-act-pair · gpt-3.5-turbo

phrases related to reading and further information

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

Ã±o

-0.71

ufact

-0.69

gger

-0.67

mble

-0.66

pload

-0.66

cream

-0.64

Ĭ±

-0.63

pter

-0.63

oux

-0.62

cffffcc

-0.62

POSITIVE LOGITS

 aloud

0.88

Write

0.84

 developments

0.80

just

0.76

 comprehension

0.75

enza

0.75

 excerpts

0.71

âĨĳ

0.70

 blogs

0.70

 Below

0.69

Activations Density 0.027%

links and references for further reading

phrases related to reading and further information

No Comments

No Known Activations