INDEX

Explanations

questions starting with "How does" or "How do"

oai_token-act-pair · gpt-3.5-turbo

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

Hug

-0.74

iken

-0.71

bsp

-0.70

tto

-0.69

gio

-0.68

hao

-0.68

Knight

-0.67

boa

-0.66

tails

-0.65

POSITIVE LOGITS

 reconcil

0.92

 reconcile

0.79

?),

0.78

?]

0.77

 coping

0.76

0.72

 reacting

0.71

?!

0.70

 reconciliation

0.69

?)

0.68

Activations Density 0.034%

questions starting with "How does" or "How do"

No Comments

No Known Activations