INDEX

Explanations

weighted terms related to geographical locations

oai_token-act-pair · gpt-3.5-turbo

occurrences of the word "Count" in various contexts

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

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

 Harbor

-0.69

 inflamm

-0.68

 obser

-0.67

 Palestin

-0.67

OPLE

-0.67

Else

-0.63

Sov

-0.62

insula

-0.59

avy

-0.59

 Condition

-0.58

POSITIVE LOGITS

enance

1.95

ess

1.17

esses

1.13

downs

1.04

ries

1.01

rified

0.87

reys

0.85

down

0.83

ere

0.82

 Dracula

0.81

Activations Density 0.024%

weighted terms related to geographical locations

occurrences of the word "Count" in various contexts

No Comments

No Known Activations