INDEX

Explanations

mentions of performance levels or states, especially in reference to sports or achievements

oai_token-act-pair · gpt-3.5-turbo

repetitions of the word "form" 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 @ 6-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.6.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.6.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

EStreamFrame

-0.79

Rare

-0.73

ILLE

-0.68

 Dish

-0.64

Priv

-0.64

âĻ¦

-0.61

 Witnesses

-0.59

Se

-0.59

ç·

-0.59

 privat

-0.59

POSITIVE LOGITS

aldehyde

1.51

ulating

1.12

idable

1.08

ative

1.05

ality

1.03

ulates

1.02

atter

1.00

ulations

0.99

ulators

0.96

atted

0.96

Activations Density 0.019%

mentions of performance levels or states, especially in reference to sports or achievements

repetitions of the word "form" in various contexts

No Comments

No Known Activations