INDEX

Explanations

words related to strong emotional reactions such as tears

oai_token-act-pair · gpt-3.5-turbo

references to emotional expressions, particularly tears

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 2-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.2.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.2.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

stood

-0.70

ancial

-0.69

offic

-0.67

ENCY

-0.64

Trend

-0.64

 Senegal

-0.62

Ü

-0.61

Notable

-0.61

enture

-0.60

pmwiki

-0.60

POSITIVE LOGITS

 tears

0.99

bows

0.98

stals

0.84

 Tears

0.84

stained

0.83

terday

0.82

fters

0.79

mith

0.77

bow

0.75

 beads

0.75

Activations Density 0.010%

words related to strong emotional reactions such as tears

references to emotional expressions, particularly tears

No Comments

No Known Activations