INDEX

Explanations

phrases related to legal or authoritative actions

oai_token-act-pair · gpt-3.5-turbo

phrases related to enforcement actions or efforts to reduce certain behaviors

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

 sake

-0.71

ensical

-0.69

æµ

-0.68

join

-0.64

 Upton

-0.63

hof

-0.61

 Norn

-0.61

 anecdote

-0.61

 Nobel

-0.60

 Tribe

-0.60

POSITIVE LOGITS

orously

0.81

 enforcement

0.76

 territ

0.76

senal

0.71

 enforced

0.71

agra

0.70

ERA

0.69

 administr

0.68

 against

0.67

 enforcing

0.67

Activations Density 0.050%

phrases related to legal or authoritative actions

phrases related to enforcement actions or efforts to reduce certain behaviors

No Comments

No Known Activations