INDEX

Explanations

instances of the word "handle" or its variations

oai_token-act-pair · gpt-3.5-turbo

phrases related to managing or dealing with situations or responsibilities

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 7-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.7.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.7.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

hemat

-0.77

esis

-0.75

ischer

-0.73

mpeg

-0.73

inet

-0.68

elong

-0.67

ode

-0.67

zyme

-0.67

urt

-0.66

owicz

-0.65

POSITIVE LOGITS

bars

1.16

 mishand

0.75

 handle

0.74

 handles

0.73

 handling

0.70

 sensitive

0.70

 handled

0.70

 paperwork

0.65

 pronouns

0.65

bar

0.64

Activations Density 0.051%

instances of the word "handle" or its variations

phrases related to managing or dealing with situations or responsibilities

No Comments

No Known Activations