INDEX

Explanations

phrases related to negative emotions or reactions, particularly disappointment

oai_token-act-pair · gpt-3.5-turbo

expressions of disappointment

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.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

ioxide

-0.82

anium

-0.77

Begin

-0.68

 vantage

-0.65

tumblr

-0.64

esm

-0.64

ãĥĻ

-0.64

clerosis

-0.63

 assum

-0.62

ilitating

-0.61

POSITIVE LOGITS

by

0.85

 that

0.75

 about

0.72

by

0.69

 Hogan

0.66

By

0.65

 with

0.64

By

0.64

 McGu

0.61

SB

0.61

Activations Density 0.111%

phrases related to negative emotions or reactions, particularly disappointment

expressions of disappointment

No Comments

No Known Activations