INDEX

Explanations

words related to emotions and feelings, specifically focusing on moods

oai_token-act-pair · gpt-3.5-turbo

references to mood and emotional states

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

ces

-0.80

ced

-0.65

advertisement

-0.63

onga

-0.61

adr

-0.61

 Deadline

-0.60

Ded

-0.60

 Douglas

-0.60

 Prosecut

-0.59

PLIC

-0.59

POSITIVE LOGITS

 swings

1.10

 mood

0.89

 Mood

0.89

 Indigo

0.87

swing

0.82

 stabil

0.75

 contag

0.75

 upbeat

0.74

 disturbance

0.73

 fluctuations

0.73

Activations Density 0.041%

words related to emotions and feelings, specifically focusing on moods

references to mood and emotional states

No Comments

No Known Activations