INDEX

Explanations

research-related terms or activities

oai_token-act-pair · gpt-3.5-turbo

references to scientific research

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

 jerk

-0.61

cakes

-0.60

emonic

-0.60

 rapp

-0.60

 chopped

-0.58

awar

-0.57

 waving

-0.57

 Blaze

-0.57

asar

-0.57

 lash

-0.57

POSITIVE LOGITS

 laboratories

0.85

 labs

0.80

 scientist

0.78

 assistants

0.77

 instit

0.76

 undertaken

0.75

 conducted

0.75

çīĪ

0.74

research

0.74

 research

0.74

Activations Density 0.048%

research-related terms or activities

references to scientific research

No Comments

No Known Activations