INDEX

Explanations

phrases or expressions related to describing actions or processes

oai_token-act-pair · gpt-3.5-turbo

the term "way" 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 @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

uster

-0.83

usters

-0.80

 livest

-0.71

encer

-0.71

 sugg

-0.69

igh

-0.65

vertisement

-0.64

utch

-0.63

Gru

-0.63

iosyncr

-0.63

POSITIVE LOGITS

fare

1.06

ward

0.96

finding

0.90

WAY

0.77

point

0.77

forward

0.75

PASS

0.68

sey

0.68

Fuck

0.68

points

0.67

Activations Density 0.057%

phrases or expressions related to describing actions or processes

the term "way" in various contexts

No Comments

No Known Activations