INDEX

Explanations

references to ongoing stories or events

oai_token-act-pair · gpt-3.5-turbo

terms related to narratives or complex stories involving difficulties

oai_token-act-pair · gpt-4o-mini

Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Negative Logits

vation

-0.84

icrobial

-0.76

 occupancy

-0.72

omet

-0.71

emouth

-0.70

licks

-0.70

umper

-0.70

haps

-0.69

obal

-0.69

aez

-0.69

POSITIVE LOGITS

 involving

1.11

 unfolding

0.98

 saga

0.98

roy

0.93

 unfold

0.90

 unfolded

0.87

 revolving

0.87

 plag

0.87

 fiasco

0.87

 debacle

0.84

Activations Density 0.253%

references to ongoing stories or events

terms related to narratives or complex stories involving difficulties

No Known Activations